Diese Präsentation wurde erfolgreich gemeldet.
Wir verwenden Ihre LinkedIn Profilangaben und Informationen zu Ihren Aktivitäten, um Anzeigen zu personalisieren und Ihnen relevantere Inhalte anzuzeigen. Sie können Ihre Anzeigeneinstellungen jederzeit ändern.

Investment Management: Art Or Science

2.562 Aufrufe

Veröffentlicht am

Veröffentlicht in: Business, Wirtschaft & Finanzen
  • Als Erste(r) kommentieren

  • Gehören Sie zu den Ersten, denen das gefällt!

Investment Management: Art Or Science

  1. 1. INVESTMENT MANAGEMENT: A SCIENCE TO TEACH OR AN ART TO LEARN? Frank J. Fabozzi, CFA Sergio M. Focardi Caroline Jonas
  2. 2. Ameritech Anonymous Robert D. Arnott Theodore R. Aronson, CFA Asahi Mutual Life Batterymarch Financial Management Boston Company Boston Partners Asset Management, L.P. Gary P. Brinson, CFA Brinson Partners, Inc. Capital Group International, Inc. Concord Capital Management Dai-Ichi Life Company Daiwa Securities Mr. and Mrs. Jeffrey Diermeier Gifford Fong Associates Investment Counsel Association of America, Inc. Jacobs Levy Equity Management John A. Gunn, CFA Jon L. Hagler Foundation Long-Term Credit Bank of Japan, Ltd. Lynch, Jones & Ryan Meiji Mutual Life Insurance Company Miller Anderson & Sherrerd, LLP John B. Neff, CFA Nikko Securities Co., Ltd. Nippon Life Insurance Company of Japan Nomura Securities Co., Ltd. Payden & Rygel Provident National Bank Frank K. Reilly, CFA Salomon Brothers Sassoon Holdings Pte. Ltd. Scudder Stevens & Clark Security Analysts Association of Japan Shaw Data Securities, Inc. Sit Investment Associates, Inc. Standish, Ayer & Wood, Inc. State Farm Insurance Companies Sumitomo Life America, Inc. T. Rowe Price Associates, Inc. Templeton Investment Counsel Inc. Travelers Insurance Co. USF&G Companies Yamaichi Securities Co., Ltd. Named Endowments The CFA Institute Research Foundation acknowledges with sincere gratitude the gen- erous contributions of the Named Endowment participants listed below. Gifts of at least US$100,000 qualify donors for membership in the Named Endowment category, which recognizes in perpetuity the commitment toward unbiased, practitioner- oriented, relevant research that these firms and individuals have expressed through their generous support of the CFA Institute Research Foundation. For more on upcoming Research Foundation publications and webcasts, please visit www.cfainstitute.org/learning/foundation/. Research Foundation monographs are online at www.cfapubs.org. Senior Research Fellows Financial Services Analyst Association The CFA Institute Research Foundation Board of Trustees 2013–2014 Chair Jeffery V. Bailey, CFA Target Corporation Manu Bhaskaran Centennial Asia Advisors Pte Limited Renee Kathleen-Doyle Blasky, CFA Vista Capital Limited Dwight Churchill, CFA Sunapee, NH William Fung London Business School Diane Garnick Clear Alternatives LLC John T. “JT” Grier, CFA Virginia Retirement System Pranay Gupta, CFA Lombard Odier Darier Hentsch (Asia) Limited Walter V. “Bud” Haslett, Jr., CFA CFA Institute George R. Hoguet, CFA, FRM State Street Global Advisors Aaron Low, CFA Lumen Advisors Alan Meder, CFA Duff & Phelps Investment Management John D. Rogers, CFA CFA Institute Brian Singer, CFA William Blair, Dynamic Allocation Strategies Wayne H. Wagner Venice Beach, CA Arnold S. Wood* Martingale Asset Management Officers and Directors Executive Director Walter V. “Bud” Haslett, Jr., CFA CFA Institute Gary P. Brinson Director of Research Laurence B. Siegel CFA Institute Research Foundation Secretary Tina Sapsara CFA Institute Treasurer Kim Maynard CFA Institute Research Foundation Review Board William J. Bernstein Efficient Frontier Advisors Stephen J. Brown New York University Sanjiv Das Santa Clara University Bernard Dumas INSEAD Stephen Figlewski New York University Gary L. Gastineau ETF Consultants, LLC William N. Goetzmann Yale School of Management Stephen A. Gorman, CFA Wellington Management Company Elizabeth R. Hilpman Barlow Partners, Inc. Paul D. Kaplan Morningstar, Inc. Robert E. Kiernan III Advanced Portfolio Management Robert W. Kopprasch, CFA The Yield Book Inc. Andrew W. Lo Massachusetts Institute of Technology Alan Marcus Boston College Paul O’Connell FDO Partners Krishna Ramaswamy University of Pennsylvania Andrew Rudd Advisor Software, Inc. Lee R. Thomas Pacific Investment Management Company Robert Trevor Macquarie University *Emeritus RF_fabozzi_cover.orig.indd Custom V 2 5/16/2014 1:32:58 PM
  3. 3. Frank J. Fabozzi, CFA Professor of Finance EDHEC Business School Sergio M. Focardi Visiting Professor of Finance Stony Brook University Caroline Jonas Managing Partner Intertek Group Investment Management: A Science to Teach or an Art to Learn?
  4. 4. Statement of Purpose The CFA Institute Research Foundation is a not-for-profit organization established to promote the development and dissemination of relevant research for investment practitioners worldwide. Neither the Research Foundation, CFA Institute, nor the publication’s editorial staff is responsible for facts and opinions presented in this pub- lication. This publication reflects the views of the author(s) and does not represent the official views of the Research Foundation or CFA Institute. The CFA Institute Research Foundation and the Research Foundation logo are trademarks owned by The CFA Institute Research Foundation. CFA®, Chartered Financial Analyst®, AIMR-PPS®, and GIPS® are just a few of the trademarks owned by CFA Institute. To view a list of CFA Institute trademarks and the Guide for the Use of CFA Institute Marks, please visit our website at www.cfainstitute.org. © 2014 The CFA Institute Research Foundation All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the copyright holder. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold with the understanding that the publisher is not engaged in rendering legal,accounting,or other professional service.If legal advice or other expert assistance is required, the services of a competent professional should be sought. ISBN 978-1-934667-73-6 16 May 2014 Editorial Staff Elizabeth Collins Book Editor Pat Light Assistant Editor Cindy Maisannes Manager, Publications Production Randy Carila Publishing Technology Specialist
  5. 5. Biographies Frank J. Fabozzi, CFA, is a professor of finance at EDHEC Business School, France, and a member of the EDHEC Risk Institute. Prior to joining EDHEC, he held various professorial positions in finance at Yale and the Massachusetts Institute of Technology. Professor Fabozzi also served as James Wei Visiting Professor in Entrepreneurship at Princeton University, where he is also currently a research fellow in the Department of Operations Research and Financial Engineering. A trustee for the BlackRock family of closed-end funds and the equity-liquidity complexes, Fabozzi has authored and edited many books on asset management. He is the recipient of the C. Stewart Sheppard Award from CFA Institute. Fabozzi received his bachelor’s and master’s degrees in econom- ics and statistics from the City College of New York and his PhD in economics from the City University of New York. Sergio M. Focardi is a visiting professor of finance at Stony Brook University, New York, and a founding partner of the Intertek Group. He serves on the edi- torial board of the Journal of Portfolio Management and has co-authored numer- ous articles and books, including the CFA Institute Research Foundation books InvestmentManagementaftertheGlobalFinancialCrisis, ChallengesinQuantitative Equity Management, and Trends in Quantitative Finance and the award-winning books Financial Modeling of the Equity Market: From CAPM to Cointegration and The Mathematics of Financial Modeling and Investment Management. Focardi also co-authored Financial Econometrics: From Basics to Advanced Modeling Techniques and Robust Portfolio Optimization and Management. He received his degree in electronic engineering from the University of Genoa and his PhD in finance from the University of Karlsruhe. Caroline Jonas is a managing partner of the Intertek Group in Paris, where she is responsible for research projects. Jonas is a co-author of numerous reports and books on finance and technology, including the CFA Institute Research Foundation books Investment Management after the Global Financial Crisis and Challenges in Quantitative Equity Management. Jonas received her bachelor’s degree from the University of Illinois at Urbana-Champaign.
  6. 6. Acknowledgments We wish to thank all those who contributed to this book, including the human resources managers at asset management firms to whom we promised anonymity. A very special thank-you goes to contributors from academia and the industry who accepted the challenge to articulate their views, based on their experiences in and observations of recent financial crises, on what might need to change in the education of future investment professionals—and indeed in the practice itself. Their views are cited and attributed throughout the book. See the section Opinion Contributors for a full list of those whose views contributed to this book. We sincerely hope that this book will contribute to the ongoing debate about what we should teach future investment professionals and, by exten- sion, have an impact on how practitioners manage other people’s money. We are also grateful to the CFA Institute Research Foundation for fund- ing this project and to its director of research, Laurence B. Siegel, for his encouragement, assistance, and insightful comments.
  7. 7. This publication qualifies for 5 CE credits, inclusive of 5 SER credits, under the guidelines of the CFA Institute Continuing Education Program. Contents Foreword................................................................................... ix 1. Finance Theory: Do We Have a Science to Teach?........................ 1 Do We Have a Science to Teach?........................................... 1 Poking Holes in the Theory................................................... 6 Completing the Theory........................................................ 10 Finance Theory as Physics Envy.............................................. 13 Finance as a Social Science?.................................................. 14 More than Simply a Social Science?........................................ 17 Finance as an Empirical Science............................................. 19 Why Are Mainstream Economic and Financial Economic Theories So Resilient?....................................................... 21 2. The Theory and Practice of Investment Management after the Crisis: Need for Change?.................................................... 27 Diversification..................................................................... 28 Optimization: Diversification Formalized................................ 35 Capital Asset Pricing Model.................................................. 39 The Efficient Market Hypothesis........................................... 42 Risk Measurement and Management.................................... 49 Crises: Do We Have the Tools for Modeling Systemic Risk?....... 56 3. Teaching Finance: Can We Do Better?........................................ 59 Is What We Are Teaching Useful?.......................................... 60 Do We Need to Change the Way We Teach Finance Theory?.... 61 Are There Any Specifics We Need to Change?......................... 62 4. What’s Missing in the Curricula for Future Investment Professionals?......................................................................... 71 Specific Topics to Reinforce or Add........................................ 73 5. How Will Future Professionals Land a Job in Investment Management?........................................................................ 91 Is There an Ideal Candidate for a Job in Investment Management?................................................................. 91 Is There a Best School?......................................................... 97 Getting through the Screening Process.................................. 97 Most Important Takeaway from Formal Education.................. 98 Opinion Contributors.................................................................. 101 References................................................................................. 105
  8. 8. ©2014 The CFA Institute Research Foundation ix Foreword Because Frank Fabozzi, Sergio Focardi, and Caroline Jonas have, in this book, looked at the question of how to teach finance from the viewpoint of instruc- tors, I will briefly consider the perspective of a student. What do I need to know? What are the timeless truths I need to understand even if there is no immediate application for them? What are the controversial propositions, and how close are we to resolving them? What is simply wrong? The basics of investment finance can be distilled down to about eight ideas: • time value of money, • discounted cash flow (as the fair value of an asset), • bond math and duration, • the no-arbitrage condition, • market efficiency, • portfolio efficiency and optimization, • the capital asset pricing model (CAPM) and market model (alpha and beta), and • option pricing and optionality. To these basics, I would add the Modigliani and Miller indifference prin- ciples for capital structure and for dividend policy; although these principles are usually taught in corporate finance rather than investment courses, they are very important for making investment decisions. That’s it. I’m done. That’s the finance course that I’d like to take—I think.i The first few ideas listed are relatively uncontroversial. But when I, as a student, get to the middle of the list, I’m tempted to howl, “Wait a minute!” Market efficiency? The market, says the great investor Jeremy Grantham, is “deliciously inefficient.” His vast fortune is testimony to the fact that somebody can beat the market. Graham and Dodd and Warren Buffett and practically every hedge fund manager would agree. So, should finance professors teach market efficiency as a timeless truth, a controversial proposition, or an idea that has been tested and found to be wrong? I would say they should teach it as a vitally important null hypothesis and point of departure for evaluating the claims of those who say they can beat the market. i That is the whole course if we are dealing with only one currency. The fact of multiple curren- cies makes finance more complicated, but international issues belong in the second semester.
  9. 9. Investment Management x ©2014 The CFA Institute Research Foundation Portfolio efficiency says that investors should try to build portfolios that maximize utility, which consists of expected return minus some measure of risk. But where are investors supposed to get their return expectations from? What is risk? Is it volatility? Downside risk? Permanent loss of capital? When I go to work in an investment management firm, will I really be building portfolios that maximize return subject to a penalty for risk, or will I be doing something else to deliver the desired results to customers? The CAPM is another problem area. The CAPM is a magnificent piece of reasoning, but the linear relationship that it posits between beta risk and expected return does not hold exactly. Active management is basically a search for assets with high returns and low risk, which the CAPM says cannot exist. The debate about the CAPM is closely related to the debate about market effi- ciency. Should professors present the CAPM as a hypothesis, as a well-reasoned framework for thinking about the relation between risk and return, or as truth? Fabozzi, Focardi, and Jonas, with whom our readers are probably already familiar from their many fine survey-based books for the CFA Institute Research Foundation, address these questions and other related issues in the current work, engagingly titled Investment Management: A Science to Teach or an Art to Learn? After interviewing finance professors, employers, and other opinion leaders in Europe, the United States, and Asia, the authors make recommendations for the teaching of finance—investment management, in particular—primarily at the MBA level. They frame their investigation in the context of the global financial crisis of 2007–2009, which caused many observers to question the basics they had been taught in finance courses. Because of CFA Institute’s origins in security analyst societies, the authors have focused on the educational needs faced by such analysts. The decision of what to teach in investment courses, however, affects the broader population now served by CFA Institute, including asset allocators, manager allocators, wealth managers, and marketing and client service professionals. Participants in all of these activities will find this book to be of great interest. The CFA Institute Research Foundation is especially pleased to present this investigation. A half century after the core of modern finance theory was developed, questioning the basic tenets of that body of work is sensible. Most of the ideas have stood the test of time, but some require revision in the light of experience. Students in our field deserve to know the best thinking of their teachers on these questions. Laurence B. Siegel Gary P. Brinson Director of Research CFA Institute Research Foundation April 2014
  10. 10. ©2014 The CFA Institute Research Foundation 1 1. Finance Theory: Do We Have a Science to Teach? In the aftermath of the 2007–09 financial crisis, mainstream finance theory was criticized for having failed to either forecast or help prevent the market crash, which resulted in large losses for investors. Although as of the writ- ing of this book at the end of 2013, markets have recovered beyond precrisis levels, the investors enjoying the recovery are not always the same investors as those who suffered the losses. So, the crash caused permanent impairment of wealth in many cases. One of the most interesting aspects of this particular crash is that finance theory, not simply the practices of the financial services industry, has been directly blamed for the crisis. That is, some observers suggest that the crash itself was the result of bad or poorly applied theory. Our goal in researching and writing this book was to explore the impli- cations of these criticisms for the curricula of finance programs at business schools and universities and, by extension, for practitioners. We begin with a discussion of finance theory as it is taught today at most institutions. In doing so, we discuss the critique and the defense of prevailing theories by integrat- ing a review of the literature and conversations with academics, asset manag- ers, and other market players. Although our focus here is finance theory, we also address economic theory to some extent because classical finance theory and classical economic theory share the same principles. Indeed, since the contribution of Eugene Fama (1965, 1970), professor of finance at the University of Chicago Booth School of Business and a corecipient of the 2013 Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel,1 the principles of neo- classical economics—in particular, the hypothesis that capital markets are efficient—have been applied to finance. Do We Have a Science to Teach? The first question is whether we have a science (or are making progress toward a science) to teach future investment professionals. Is our “science” merely an idealized rational construction that ignores market realities? If so, exactly what should we be teaching students of finance whose objective is to manage 1 Hereafter called the “Nobel Prize in Economics.” By spelling out the full name of the prize, we acknowledge that it was not in the list of prizes established by Alfred Nobel himself, but the Nobel Foundation clearly expresses its view that it is to be considered on an equal footing with the original Nobel prizes.
  11. 11. Investment Management 2 ©2014 The CFA Institute Research Foundation other people’s money? Is an alternative science based on observations available (or in progress)? Or does our current knowledge of economics and finance have to be removed from the realm of science altogether and placed on a par with the social sciences? In response to the criticisms leveled at mainstream finance theory follow- ing recurrent financial crises, the proponents of the theory defend its valid- ity. They argue that all sciences use idealizations and that the idealizations used in mainstream economics and financial economics are useful, although they cannot foresee—or explain—financial crises such as the 2007–09 crash. According to mainstream theory, the cause of large market swings is attrib- utable to exogenous events that the theory cannot predict. Others consider crashes to be the consequence of random fluctuations in market returns. This view deserves explanation. The fact that a phe- nomenon can be described with simple probabilistic models does not per se preclude the existence of a deeper, more informative explanation of the same phenomenon. Different levels of explanation might coexist, of course, with different levels of accuracy. For example, random-number generators are perfectly deterministic models that generate sequences of numbers that appear to be random sequences. Finite sequences of numbers generated by random-number generators pass all tests of randomness and are described as sequences of independent draws from a given distribution. Although these sequences are generated by a deterministic model, they can be described with good approximation as random sequences. In both the practice and the theory of finance, different families of statistical models of varying complexity can be used to describe the same data samples. The choice between these models is often based on statistical tests that do not allow any definitive answer. The possibility of describing crashes as random phenomena is not in contradiction with more refined models that have greater predictive power. By adopting appropriate distributions, one can take the simplified view that crashes are purely random events. This approach is the first level of approxima- tion, the most coarse-grained view of market behavior. The theoretical challenge, however, is to find more informative explanations—in par- ticular, explanations in which the conditional probability of market crashes depends on observed variables. This type of explanation is what is required from a theory of market crashes. In his article “In Defence of the Dismal Science” (2009), which appeared on the Economist website on 6 August 2009, Robert Lucas, pro- fessor of economics at the University of Chicago and recipient of the 1995 Nobel Prize in Economics, wrote, “One thing we are not going to have, now or ever, is a set of models that forecast sudden falls in the value of financial
  12. 12. Finance Theory ©2014 The CFA Institute Research Foundation 3 assets, like the declines that followed the failure of Lehman Brothers in September [2008].” This statement is somewhat misleading: It should be obvious that we are not going to have a deterministic model that predicts with certainty large market swings, their amplitude, and their timing. Rather what is expected of a scientific theory is that it allow to evaluate with reasonable accuracy the likelihood of a crisis. In a glib dismal of the importance of the market crash, Robert Barro (2009), professor of economics at Harvard University, remarked during a roundtable discussion published two days later on the Economist website, “Economies have natural tendencies to recover from recessions, and such a recovery is the most likely outcome for the American economy going into 2010.” In our review of the literature following the 2007–09 financial crisis and in our conversations about the topic, one of the problems singled out with the prevailing theory as presently taught in most finance curricula is that the idealizations made by mainstream finance theory fail to take into account how real-world markets work. Mainstream academics are widely considered to be more interested in the quest for a unified theory than in understanding the workings of markets. For example, in the equity mar- ket, while mainstream academics often hold that stocks are priced cor- rectly, there are, according to Dennis Logue, professor emeritus at the Tuck School of Business Administration at Dartmouth College and chairman of the board of directors of Ledyard Financial Group, “massive anomalies in the micro and macro sense.” Before discussing in more detail the defense and the critique of main- stream finance theory, we wish to briefly state what we mean by “main- stream” (or prevailing or dominant) because the term is subject to various interpretations. We use the term mainstream as shorthand for referring to the theory that is espoused in articles that appear in major journals and that is taught at major universities and business schools. We do not mean to sug- gest that every academic who might personally be considered mainstream adheres exactly to these views. The chief tenets of mainstream theory are (1) efficient markets, (2) rational expectations, and (3) optimization. In the 1961–66 period, Jack Treynor, William Sharpe, John Lintner, and Jan Mossin independently introduced the first general equilibrium theory in finance, called the capital asset pricing model (CAPM). According to the CAPM, all agents share the same knowledge of the probability distri- butions of future returns and rely on mean–variance optimization to make their investment decisions. That is, all agents choose the optimal compromise between the expected return and the expected variance of their portfolio. As
  13. 13. Investment Management 4 ©2014 The CFA Institute Research Foundation a result, they all invest in the same risky portfolio, the market portfolio. Their portfolios differ only in the amount allocated to cash (the “riskless” asset). Robert Merton (1973), distinguished professor of finance at Sloan School of Management at the Massachusetts Institute of Technology (MIT) and a corecipient of the 1997 Nobel Prize in Economics, extended the CAPM in a dynamic environment in his seminal work. The Merton model is a multi-​ period model in which decisions are made by considering not only next-period returns but also the entire future price process of assets. Mainstream economic theory developed in parallel with mainstream finance theory in the 1960s and 1970s in what is called the “rational expec- tations revolution.” The starting point was the so-called Lucas critique. Professor Lucas observed that the estimation of the effect of changes in gov- ernment policy is made ineffective by the fact that economic agents anticipate these changes and change their behavior. Therefore, he advocated giving a micro foundation to macroeconomics—that is, explaining macroeconomics in terms of the behavior of individual agents. The result was a tendency in mainstream economic theory for macro-​ economic models to be based on a multitude of agents characterized by ratio- nal expectations, optimization, and equilibrium. Mainstream finance theory uses the same basic structure as general equilibrium economics. It assumes markets are populated by a multitude of agents and each agent is identified by a utility function that assigns a numerical value to each possible invest- ment choice. Each agent receives a stochastic (i.e., random) stream of endow- ments (i.e., exogenous positive cash flows). Endowments can represent any cash flow received outside of financial investments, such as salaries, gifts, or inheritances. At each trading moment, agents decide how much they want to consume, how much they want to invest in financial assets, and how much they want to keep as cash. The principle of dynamic equilibrium in finance theory requires that at each moment, prices are such that the global demand for assets is equal to the global offer of assets. In the absence of arbitrage, the assumption is that all agents can be aggregated into a single representative agent. The consumption stream and the price process generated by this representative agent are the same as the aggregated consumption and relative price processes obtained by optimizing individual agents. The assumptions made in mainstream finance theory are clearly unre- alistic. So, is mainstream finance theory (or, generally, current mainstream macro​economic theory) an empirical science at all in the modern sense? That is, is the theory based on observations? Many would argue that financial economics does not belong to the realm of empirical science but to that of the social sciences. Michael Oliver, a senior
  14. 14. Finance Theory ©2014 The CFA Institute Research Foundation 5 lecturer in finance at the Open University and cofounder and director of Global Partnership Family Offices, remarked, “Economics is a social science, not a physical science.” The meaning behind this remark is that separating pure economics from political economics is difficult. In short, different economic theories corre- spond to different political choices. Economics and finance have as their sub- ject an artifact, the economy or the markets, not laws of nature. The artifact is context specific: It is not independent of social or political objectives. Hence, separating empirical laws from statements of principles is not easy. In his article “How Should the Financial Crisis Change How We Teach Economics?” (2010a), Robert Shiller, professor of economics at Yale University and a corecipient of the 2013 Nobel Prize in Economics, remarked on the number of critics of current mainstream economics. He concluded, “The reason there are such strong views about the profession going astray is that we do not have good scientific macroeconomic theories; we do not even have good ways of developing them” (p. 406). Some have argued that the reason mainstream macroeconomics and mainstream finance theory are not scientific can be found in the design of these disciplines. John Kay, a distinguished British economist and visit- ing professor at the London School of Economics, observed that main- stream economics is a logical theory based on unrealistic assumptions without any consideration of real data. Professor Kay (2012) observed, “The distinguishing characteristic of [mainstream economists’] approach is that the list of unrealistic simplifying assumptions is extremely long” (p. 50). Discussing the ineffectiveness of policy—and, we might add, invest- ment decisions—based on the assumptions of modern macroeconomics, Professor Kay went on to cite John Cochrane, professor of finance at the University of Chicago’s Booth School of Business, who agrees that the assumptions used “are, as usual, obviously not true” (p. 51). That, Professor Kay remarked, would be the end of the discussion for any reasonable “sci- entist.” Professor Cochrane argued, however, that “this [endlessly playing with unrealistic hypotheses] is exactly the right way of doing things.” In the same article, Professor Kay commented on the absurdity that a priori deduction from a particular set of unrealistic simplifying assumptions is not simply a tool but, as stated by the University of Chicago’s Gary Becker, winner of the 1992 Nobel Prize in Economics, “the heart of the economic approach” (p. 55). Exhibit 1.1 summarizes the defense and some of the critiques of main- stream economic and finance theory and notes some elements that have been proposed that would characterize an alternative theory.
  15. 15. Investment Management 6 ©2014 The CFA Institute Research Foundation Poking Holes in the Theory Mainstream finance theory is considered to be unrealistic not only because its main assumptions are unrealistic but also because the entire theoreti- cal construction is not related to observable quantities. For example, such crucial data as future dividends and returns are not observable. In his book Dynamics of Markets (2009), University of Houston professor of physics Joseph McCauley noted, The idea of dividends and returns discounted infinitely into the future for financial assets is very shaky, because it makes impossible information demands on our knowledge of future dividends and returns. That is, it is impossible to apply with any reasonable degree of accuracy. (p. 65) The fact that the theory makes impossible demands on our knowledge is a crucial point that affects all mainstream general equilibrium theories. Fundamental theoretical variables, such as prices, are defined as the dis- counted present value of an infinite stream of future quantities that are not observable. Contrast this circumstance with physics, in which many theoreti- cal terms are not directly observable but are defined through the theory itself. Consider temperature: We cannot directly observe temperature, which is a theoretical term interpreted as the amount of energy associated with the motion of certain molecules. All theoretical terms used to define temperature, however, are defined in function of observables. For exam- ple, suppose you measure the temperature of the body by using a clinical thermometer with a mercury column. What you actually observe is not temperature but the elongation of the mercury column. We translate the elongation of the mercury column into temperature because we have a global theory that links temperature with other observable characteristics Exhibit 1.1.  Defense and Critiques of Mainstream Economic and Finance Theory and Alternatives Defense of Mainstream Finance Theory Critique of Mainstream Finance Theory Elements for an Alternative Theory Mainstream finance theory is an idealized but valid representation of financial markets. Crises are unpredictable events and are subsequently self-correcting. Mainstream finance theory models of rationality, agent independence, and equilibrium are unrealistic. Markets are neither stable nor self-regulating as held by equilibrium assumptions. Markets are complex systems based on interacting (noncollapsible and not necessarily rational) agents. Markets are prone to crises because of aggregation phenomena. The money generation process is an essential component that leads to bubbles and crashes.
  16. 16. Finance Theory ©2014 The CFA Institute Research Foundation 7 such as length and volume. These terms are, indeed, observable. Thus, temperature can be defined, and it is a useful concept because it helps explain other observed phenomena. Economic and finance theory, on the contrary, define terms in function of quantities that are not observable, nor can they be defined in function of observables. Quantities such as future dividends are not defined through a process of forecasting based on past data. If these terms were defined as a function of past data, then mainstream finance would be based on observ- able data. Mainstream finance, however, is based on future, clearly non- observable, data. In practice, any present value model of asset prices—that is, any model that says that today’s price is based on discounted future cash flows—makes forecasts of unobservable future quantities. In addition to this problem, which is fundamental, the critique of mainstream finance theory makes three key points that can be summarized as follows: 1. No real agent has a perfect knowledge of the future, not even in a proba- bilistic sense. Hence, the notion of rational expectations is unrealistic. 2. Because real agents have mutual interactions and are not coordinated solely by a central price signal, agents cannot be collapsed into a single representative agent.2 3. Economies are rarely in a state of equilibrium. Alan Kirman (2009), professor emeritus of economics at the University of Aix-Marseille III and at the École des Hautes Études en Sciences Sociales, remarked, What has become the standard macroeconomic model . . . is justified by its proponents on the grounds that it . . . is based on rational maximis- ing individuals. But there are two problems with this. . . . First, we have known since the mid-1970s that aggregating the behaviour of lots of ratio- nal individuals will not necessarily lead to behaviour consistent with that of some “representative agent”. . . . Second, the axioms that are used to define “rationality” are based on the introspection of economists and not on the observed behaviour of individuals. (pp. 80–81) 2 Neoclassical economics does not posit or require a representative agent but, instead, sup- poses that different agents will have different utility functions and that the market-clearing price will represent the net effects of all the agents in the market. In contrast, much of modern macroeconomics relies on a single representative agent. Without the assumption of a representative agent, the dynamic stochastic general equilibrium models most often used in macroeconomics are neither mathematically nor computationally tractable. There is no way to solve a dynamic stochastic optimization problem with a large number of independent utility functions.
  17. 17. Investment Management 8 ©2014 The CFA Institute Research Foundation How unrealistic are rational expectations? Eric Beinhocker (2007), exec- utive director of the Institute for New Economic Thinking’s research program at the University of Oxford (INET@Oxford), asked the reader to consider a rational agent who goes grocery shopping:3 You have well-defined preferences for tomatoes compared with every- thing else you could possibly buy in the world, including bread, milk, and a vacation in Spain. Furthermore, you have well-defined preferences for everything you could possibly buy at any point in the future, and since the future is uncertain, you have assigned probabilities to those poten- tial purchases. For example, I believe that there is a 23% chance that in two years, the shelf in my kitchen will come loose and I will need to pay $1.20 to buy some bolts to fix it. The discounted present value of that $1.20 is about $1.00, multiplied by a 23% probability, equals an expected value of 23 cents for possible future repairs, which I must trade off with my potential purchase of tomatoes today, along with all of my other potential purchases in my lifetime. . . . [To make your decisions,] you know exactly what your budget is for spending on tomatoes. To calculate this budget, you must have fully formed expectations of your future earnings over your entire lifetime and have optimized your cur- rent budget on the basis of that knowledge. In other words, you might hold back on those tomatoes because you know that the money spent on them could be better spent in your retirement. Of course, this assumes that your future earnings will be invested in a perfectly hedged portfolio of financial assets and that you take into account actuarial calculations on the probability that you will live until retirement at age 65, as well as your expectations of future interest rates, inflation, and the yen-to-dollar exchange rate. While standing there, staring at those nice, red tomatoes, you then feed all this information into your mind and perform a cunning and incredibly complex optimization calculation that trades off all these factors, and you come up with the perfectly optimal answer—to buy or not to buy! (p. 116) This description might look like a caricature, but it is exactly what is implied by a rational expectations model. According to the view of positive economics, mathematical models describe the outcome of financial decisions, not the process itself. This view, which says that the aggregate supply and demand is determined “as if” all these calculations took place, weakens the Lucas critique, which calls for a 3 The Institute for New Economic Thinking (INET) is a not-for-profit think tank whose purpose is to support academic research and teaching in economics “outside the dominant paradigms of efficient markets and rational expectations.” Founded in 2009 with the financial support of George Soros, INET is a response to the global financial crisis that started in 2007. For more information, see http://ineteconomics.org/.
  18. 18. Finance Theory ©2014 The CFA Institute Research Foundation 9 microstructure foundation to macroeconomics, and is basically beyond any reasonable empirical test. As for the second critique—that agents cannot be collapsed into a sin- gle representative agent—the Sonnenschein–Debreu–Mantel theorem (see Sonnenschein 1972) demonstrated that utility functions cannot be aggregated into the utility function of a single representative agent. The idea that agents have mutual interactions and are not coordinated solely by a central price signal was analyzed two decades ago by Professor Kirman (1992). Kirman (2010) subsequently wrote, [Macroeconomics is based on the assumption that] all that we have to do to deduce the behaviour of the economy at the aggregate, or macro, level is to add up the behaviour of the individuals who make it up. Furthermore, the theoretically unjustified assumption is made that the behaviour of the aggregate can be assimilated to that of an individual. (p. 501) The critique that the representative agent is not a sound concept is based on the fact that one cannot aggregate utility functions and obtain a utility func- tion with all the characteristics needed to justify equilibrium. Agents inter- act directly, for example, in herding behavior, as is well documented in the behavioral finance literature. Paul Ormerod and Dirk Helbing (2012) wrote, We live now in a densely networked, strongly coupled, and largely inter- dependent world, which behaves completely differently from a system of independently optimizing decision makers. . . . The representative agent approach must be abandoned. . . . [It] cannot describe cascading effects well. These are determined not by the average stability, but by the weakest link. (p. 149) As for the third critique—that markets are rarely in a state of equilibrium—critics of mainstream economic and finance theory point to the frequency and the magnitude of financial crises. At the 2013 International Monetary Fund (IMF) global economy forum, David Romer (2013), professor of political economy at University of California, Berkeley, remarked, “My view that we should think of financial shocks as closer to commonplace than to exceptional is based on history.” Professor Romer counted six distinct shocks in US markets during the past 30 or so years that have posed important mac- roeconomic risks. Joseph Stiglitz (2013), professor of economics and University Professor at Columbia University and a corecipient of the 2001 Nobel Prize in Economics, counted approximately 100 financial crises worldwide in the past 30 years. Following closely on the 1987 stock market crash and 2000–01 bursting of the dot-com bubble, the most recent crisis has made it clear that
  19. 19. Investment Management 10 ©2014 The CFA Institute Research Foundation tensions accumulate in economies and markets that lead to disequilibria and large market swings. Completing the Theory Mainstream economics and mainstream economists fail to recognize the exis- tence of bubbles. In an interview, New Yorker columnist John Cassidy (2010) questioned Eugene Fama about efficient markets and the recent credit bubble in the US housing market. Professor Fama famously replied, “I don’t know what a credit bubble means. I don’t even know what a bubble means. These words have become popular. I don’t think they have any meaning.” Nevertheless, attempts have been made to explain market bubbles and crashes within (or alongside) the existing theory. Among these are attempts to integrate into finance the consideration of liquidity, leverage, and other factors outside clas- sical financial theory and to incorporate psychology (human behavior). The Open University’s Dr. Oliver commented on the importance of liquidity in explaining large stock market swings. He said, Until the financial crisis, the role of money was not taken seriously by most economists. Some economics models of the economy were even constructed without a banking system! The role of money (the term used by practitioners is “liquidity”) needs to be reassessed. Dr. Oliver collaborated with Gordon Pepper on the book The Liquidity Theory of Asset Prices (2006) and teaches the unit on liquidity during a two- day course titled “A Practical History of Financial Markets” at Edinburgh Business School. The role of liquidity in the formation of sharp upward and downward market swings is now widely recognized, but will that recognition be enough to complete mainstream finance theory? Some sources we talked to are either not convinced that incorporating liquidity in asset-pricing models would improve our theory or models or consider it too early to tell. Sébastien Lleo, professor of finance at NEOMA Business School4 (France) and visiting pro- fessor at the Frankfurt School of Finance and Management, cautioned, “We should be wary of claims that a single theory or tool can ‘fix’ our approach to finance. This will take a long time and require significant efforts.” A longer list of what is needed to rethink finance theory to take into con- sideration the real world was suggested by James Montier, a strategist with fund manager GMO. In his Manifesto for Change in his white paper “The Flaws of Finance” (2012), Mr. Montier suggested incorporating (together with liquidity) leverage, bad behavior, bad incentives, and delegated management. 4 NEOMA Business School was formed by the recent merger of Rouen Business School and Reims Management School.
  20. 20. Finance Theory ©2014 The CFA Institute Research Foundation 11 The role of human behavior in explaining large market swings has been explored by, among others, Professor Shiller. In his recent article “Bubbles Forever” (2013) on Project Syndicate, Professor Shiller suggested that bubbles might best be referred to as speculative epidemics: Enthusiasm spreads from person to person and, in the process, amplifies stories that might justify asset price increases. Shiller explored how psychological factors drive stock markets in his book Irrational Exuberance, first published in 2000 and updated in 2005. Andrew Lo (2004), professor of finance and director of the Laboratory for Financial Engineering at MIT, developed what he calls the “adaptive market hypothesis.” He argues that markets are not static but that they evolve continu- ously, not only under the pressure of exogenous events but also because of the competitive action of market participants. Professor Lo suggests that by apply- ing the principles of evolution (competition, adaptation, and natural selection) to financial markets, we can explain the behavior of markets. In fact, he compares markets to ecologies competing for resources (i.e., profits). Market participants learn from experience and modify their forecasts and investment strategies to realize a gain. In competing for resources, the action of market participants tends to keep markets efficient while creating new opportunities for profit. Note that, together with Lars Peter Hansen, professor of economics at the University of Chicago and a corecipient of the 2013 Nobel Prize in Economics, Professor Lo codirects the Macro Financial Modeling Group at the Becker Friedman Institute. The group consists of a network of macro-​ economists working to develop improved models of the links between finan- cial markets and the real economy in the wake of the 2007–09 financial crisis—a link that sources mentioned is lacking in today’s theory. One attempt to establish a historical link between the economy and mar- kets (and predict the next growth cycle) was recently made by Hans-Joerg Naumer, head of capital markets and thematic research at Allianz Global Investors. Using the Russian economist Nikolai Kondratiev’s theory of long waves of boom–bust business cycles and stock market data from Robert Shiller’s Irrational Exuberance (2005) and Datastream, Mr. Naumer overlaid a rolling 10-year yield on the SP 500 Index on Kondratiev’s five long waves (see Figure 1.1).5 Mr. Naumer’s link is of an economic nature; that is, it associates long-term stock market trends with long business cycles. This link is differ- ent from the cycles implied by Minsky’s financial instability hypothesis, which links the economy, financial markets, and the money generation process. 5 Nikolai Dmitriyevich Kondratiev (or Kondratieff) was a Russian economist who lived from 1892 to 1938 and was known for his theory that Western capitalist economies have long-term (50–60 year) cycles characterized by successions of expansion and decline. These cycles are known as “Kondratiev waves.” Kondratiev developed the theory in his 1925 book The Major Economic Cycles.
  21. 21. Investment Management 12 ©2014 The CFA Institute Research Foundation Figure 1.1.  Kondratiev’s Five Waves from 1780 to 2010 and the Rolling 10-Year Yield on the SP 500 KondratievWaveRolling10-YearYieldonSP500 Yield(%) 18 16 14 12 10 8 6 4 2 0 –2 –4 –6 –8 –10 1821518141316171911901112131415161718191200110 1stKondratiev 1780–1830 Steamengine, Textileindustry 2ndKondratiev 1830–1880 Railway, Steel 3rdKondratiev 1880–1930 Electrification, Chemicals 4thKondratiev1930–1970 Automobiles, Petrochemicals Panicof1837 1837–1843 LongDepression 1873–1879 GreatDepression 1929–1939 1st2nd Oilcrisis 1974–1980 FinancialCrisis 2008–2011 5thKondratiev 1970–2010 Information technology 6th Kondratiev 2010– 20XX ??? Source: Naumer/Allianz Global Investors (2013).
  22. 22. Finance Theory ©2014 The CFA Institute Research Foundation 13 Finance Theory as Physics Envy One might ask: Can the debate on the tenability of today’s finance theory be resolved with the methods of empirical science? Will the debate remain at the level of dogma, as with the conflict between different views of political economics? Or will the debate remain at the epistemological level, centered on the question of what is the cognitive value of a model that, in the best case, captures only some general features of the real economy and real markets? As mentioned previously, Lucas maintains that we will never have a set of models that forecasts sudden falls in the value of financial assets. He is refer- ring to sure deterministic predictions. But mainstream economic and finance theories do make probabilistic predictions. The problem is that testing predic- tions is difficult when samples are small and noise abounds. In his famous paper “Noise,” the late Fischer Black (1986) wrote, “. . . noise makes it very dif- ficult to test either practical or academic theories about the way that financial or economic markets work. We are forced to act largely in the dark” (p. 529). Do we have a science? Would you feel safe flying if you knew that there were linear differential equations that describe an airplane’s structure but that no such equations can be identified? The abstract mathematical knowledge that structures can be described by linear differential equations allows one to neither engineer nor study any real structure. Yet, this knowledge is the knowledge embedded in general equilibrium models. One objection to this critique is that we can have an understanding of economics that cannot be formalized in a mathematical model. This objec- tion is likely to be true—the Wright brothers, who were bicycle mechanics, designed their planes “as if” they had the mathematical knowledge of the structure—but the objection does not lend any support to mainstream mod- els. If we can describe economic behavior without models, we do not need general equilibrium theories. Ultimately, the debate on general equilibrium models in economics and finance theory may be empty. Clearly, general equilibrium models are not empirically validated in terms of the characteristics and interactions of real agents. Given any asset-pricing model that does not admit arbitrage, however, we can always formulate an equivalent abstract general equilibrium model. In classical physics, the laws of motion can be expressed either through differential equations or through the minimization of a functional, the Hamiltonian or the Lagrangian.6 The predictive power of physics depends on the fact that we know how to write Hamiltonian and Lagrangian terms. The mere existence of a Hamiltonian functional does not, however, add to our understanding of a physical phenomenon. 6 This statement refers to a purely mathematical fact. The differential equations of dynamics can be obtained as the solution of a maximization problem.
  23. 23. Investment Management 14 ©2014 The CFA Institute Research Foundation In finance theory, we do not know how to describe a representative agent based on empirical data, nor can we empirically ascertain the functional form of a representative agent for large markets. The pure mathematical existence of an abstract mathematical representative agent does not add much to our economic understanding of financial markets. Consider the simplest general equilibrium model, the capital asset pric- ing model (CAPM). Given a set of expected returns, we can always think of these expected returns as generated by the CAPM. This pure mathematical abstraction is always true. Of course, real agents do not behave as prescribed by the CAPM. In addition, if we go beyond a single period, which is the time horizon of the CAPM, then its predictions are no longer valid. We can always find a dynamic version of a general equilibrium model, however, that can generate any stream of returns. The problem is that we have no way to actually estimate such a model from empirical data. We explore the implications of these ideas on the teaching of finance in Chapter 3. As for the theory and the actual practice in investment management in the postcrisis period, Jaap van Dam, head of strategy and research at the Dutch pension fund PGGM (with more than €131 billion in assets under management), remarked, More than in changing the [prevailing] tenets themselves, their application in investment management is changing and they are being complemented with empirical analysis and common sense. What we need to reconsider is the universal applicability of these tenets and to admit their inherent limita- tions. A theory is just a theory. A typical formulation of a theory is of the type “if X, then Y.” Understanding the limitations of the “if X” part has probably become more important. This applies to theories like CAPM, for example, which is now best viewed as an idealized model. Commenting on market equilibria and typical no-arbitrage assump- tions, Steven Greiner, director of portfolio risk at FactSet Research Systems, remarked, “[These] are not so relevant as professors think for the practice of asset management. It is enough to know that efficiency rises with liquidity and that mispricing is empirically demonstrable.” Finance as a Social Science? If prevailing theory indeed fails to represent the world as it is and has effectively proved to be of little practical use, can we consider our economic and finance the- ory to be hard science? Wouldn’t it be better to reinstate economics and finance as social sciences, albeit quantitative social sciences (given the inherently quan- titative nature of the data), and allot a reduced role to the complex mathematics and modeling (in light of the problems with the theory behind the math)?
  24. 24. Finance Theory ©2014 The CFA Institute Research Foundation 15 Dr. Oliver remarked, Over the past 20 years I have watched in despair as universities and business schools have grilled students with existence theorems and trained them to be competent as mathematicians, frequently at the expense of understand- ing how the real-world macroeconomy works. Two arguments can be raised against considering economic/finance theory to be a mathematical science. The first is that economics and finance are dominated by single events that cannot be predicted or even described in mathematical terms. Nassim Taleb, professor of risk engineering at Polytechnic Institute of New York University and author of The Black Swan (2010), advocates this view. He popularized the notion of “black swans,” unpredictable events that change the course of an economy and that are wrongly rationalized after they occur. The key question is not whether unpredictable events occur. Of course, they do. In corporate finance, some decisions made by senior managers are difficult to model. In political economics, some key decisions made by heads of states or central banks are difficult to predict. Changes in the behav- ior of masses—such as herding, which changes the demand for an entire market—are also difficult to predict. The crucial question is whether these events can be handled with statistical techniques or whether the complex- ity of the economic system makes individual events critical for the future development of an economy or markets and thus not susceptible to statisti- cal treatment. The second argument in favor of considering finance to be more a social science than a physical science is that the dynamics of economic and finan- cial phenomena are simply too complex to be captured by mathematical formulas—at least with today’s mathematics. Or perhaps the phenomena are too complex to allow a parsimonious mathematical description. But this characteristic, the proponents of a reduced role for mathematics argue, does not imply that we cannot make empirically meaningful economic state- ments outside a mathematical model. This camp observes that economic thinking existed well before the mathematization of economics and finance. Basic economic ideas can be explained in plain English, and reasoning on economic and financial facts can be done without formulas. Russell Napier, a consultant with CLSA Asia-Pacific Markets and author of The Anatomy of the Bear: Lessons from Wall Street’s 4 Great Bottoms (2005) argued, Finance is all about establishing value. To do so, we need a better under- standing of humans, we need to remove finance from the field of science and place it more in sociology. Sociology today cannot be used as a predictive force but a field for learning. Sometime in the future, finance might migrate back to being a science. But, we cannot afford to have more theoretical culs
  25. 25. Investment Management 16 ©2014 The CFA Institute Research Foundation de sac. We cannot afford more problems deriving from the spurious certain- ties often inherent in the pricing of derivatives. Similar views have been expressed by others. For example, British econo- mist John Kay (2012) wrote, Economic behavior is influenced by technologies and cultures, which evolve in ways that are certainly not random but which cannot be described fully, or perhaps at all, by the kinds of variables and equations with which econo- mists are familiar. (p. 52) Whether we view our economic and finance theory as a hard science or as a social science influences what we teach, which we explore in Chapter 3. Commenting on the present-day emphasis on mathematics in finance programs, Dr. Oliver remarked, Many of the recently introduced programs at business schools and universi- ties with a concentration in mathematical finance are divorced from events in the real world. We are producing economists who can give you an equa- tion for everything but who lack any broader knowledge. Economics is a social science and not a physical science, and as such, it needs to refocus on core social science values. Even proponents of the use of models in investment management caution about their use. Professor Lleo remarked, For me the problem is not the application of mathematics in areas where we do not have a strong theory. This is rather healthy: We need models— mathematical, philosophical, sociological—to act as frames of reference if we are to tackle any significant question. The real problem is the application of mathematics in areas where we do have a strong theory. Our financial economic theory makes strong assumptions to derive strong results. The problem is that these assumptions are often unrealistic. However, we often lose sight of this fact because of the appeal and apparent universality of the “strong theory” we have developed. The existence of a strong prescriptive or normative theory necessarily generates overconfidence and leads to the application of the wrong type of conceptual tool, be it mathematical or sociological. Professor Lleo cited as an example the pricing of collateralized debt obli- gations (CDOs), for which, he said, we do have a strong theory (no-arbitrage pricing via hedging/replication) that enabled us in the past to use advanced mathematics confidently. Unfortunately, he added, “The structure and nature of CDOs did not satisfy the fundamental assumptions, which led to disaster.” Can we have meaningful empirical knowledge even when mathemati- cal modeling is not possible? The answer is clearly yes. For example, we can describe fairly well, in plain English, the process of the growth of a tree even
  26. 26. Finance Theory ©2014 The CFA Institute Research Foundation 17 if we do not have a detailed mathematical description of the growth of trees. Generally, we can say that many levels of description of phenomena are pos- sible. We have many levels of “coarse graining” in mathematical descriptions, but in addition, we have descriptions in natural languages that, although less precise than mathematical descriptions, are still meaningful. Forcing mathematization can actually impoverish, not enrich, knowl- edge. The imposition of a mathematical language may make important facts impossible to convey. Professor Lleo believes that economic thinking became poorer in some aspects just as it was becoming more structured and precise in others. He cites the work of Frank Knight (1921), who introduced the distinc- tion between risk and uncertainty, and of John Maynard Keynes (1936), who introduced the notion of “animal spirits.” “Yet,” Professor Lleo commented, “finance theory tells us a different story: Uncertainty can be viewed as idio- syncratic risk and diversified away. The only source of return should be related to market risk premia and the scaling of risk exposure.” More than Simply a Social Science? Although some argue that economics and finance should be considered social sciences, others argue for stricter adherence to the paradigm of empirical sci- ence. Again, the impact on the curriculum is not negligible. The discussion of the role of mathematics in scientific enquiry is not new: The entire development of science was marked by a debate on the use of mathematics. Galileo Galilei was the first to state that science was inherently mathematical. In his The Assayer, Galileo (1623) wrote, [The Book of Nature] . . . is written in the language of mathematics, and its characters are triangles, circles, and other geometrical figures, without which it is humanly impossible to understand a single word of it; without these, one is wandering around in a dark labyrinth. This statement was prophetical but in advance of its time: With the mathe- matics known to Galileo, one could not have formulated modern physics. Only later, with the development of calculus by Gottfried Leibnitz and Sir Isaac Newton, did mathematics acquire the tools for formulating mechanical laws in mathematical terms. The publication of Newton’s Principia Mathematica in 1687 marked the beginning of modern mathematical science. The mathematics of calculus—in particular, differential equations that link variables with their rate of change—proved to be a powerful concept in all scientific disciplines. Still, prior to the invention of computers, the practical application of mathematics was limited to establishing general properties, such as the exis- tence of solutions of differential equations, and finding closed-form solutions of differential equations. Thus, many problems in empirical science were not formalized mathematically. For example, empirical problems related to
  27. 27. Investment Management 18 ©2014 The CFA Institute Research Foundation weather forecasts, biology, botany, hydrology, even the design of mechanical structures, were not fully formalized. Nevertheless, these problems are part of empirical science. Quantitative laws did apply, but they were far from provid- ing a full mathematical description of these phenomena. Often, the solution to engineering problems continued to require human judgment. The introduction of high-performance computers marked a new epoch in the application of mathematics to science and engineering and ushered in the application of computational mathematics. Fast computers allowed the simulation of phenomena. Instead of being limited to closed-form solutions of differential equations, analysts have been able to actually create, through simulation, structures of numbers or symbols that mimic the structure of reality. This ability greatly increased the number of areas for the practical applicability of mathematics. Today, we can simulate with amazing accuracy the behavior of large-scale objects, such as airplanes, or natural events, such as tornados, and possibly, because the use of mathematics in science is subject to evolution, reproduce some human cognitive functions. Many complex phenomena, however, remain beyond the ability of detailed mathematical representation, and for various reasons—including chaos and sensitivity to initial conditions, objective complexity (the extent to which the phenomenon is close to randomness), and because we do not know the laws. But these are moving targets. For example, because of improved computers and software, weather forecasting has become progressively more accurate, but as we all have noticed, it can still be wide of the mark. Professor Logue compared our ability to forecast using our economic and finance the- ory with the ability of a meteorologist. He remarked, Our inability to forecast is a “super problem.” As with the weather system, it is very difficult to identify where we are now and even more difficult to identify where we will be in the future. We have all heard the local radio weatherman say 60% chance of rain while at the same time looking out the window at a deluge. David Colander, professor of economics at Middlebury College, Vermont, gave the argument a twist. He remarked (2009), “The problem is not that eco- nomics is too mathematical; it is that the mathematics we use in economics is way too simple to capture the complexities of economic interrelationships” (p. 12). Others agree and have argued that this situation calls for greater use of reasoning in managing assets. Edward Qian, chief investment officer (CIO) and head of multiasset research at PanAgora, said that the ability to reason on issues in finance and economics is what is critical; mathematics provides a tool for reasoning. He commented, Finance is based on powerful ideas and insights about the market, it is not based on powerful mathematics. But as the field evolves, it seems it has
  28. 28. Finance Theory ©2014 The CFA Institute Research Foundation 19 shifted to more mathematics and more sophisticated models. In doing so, it is easy to forget the underlying assumptions, some of which are highly unre- alistic. In recent years, students with a strong mathematics and computer science background, who would have gone to mathematics and science pro- grams, are recruited to finance and economic graduate programs. But only those who can think deeply and independently about issues in finance and economics can be expected to become successful investors. Clearly, in some domains of empirical science, all-encompassing math- ematical formulations are not possible. Economics and financial economics are probably only partially susceptible to mathematical theories. Although mathematical reasoning is useful, it probably has to be complemented with less formal reasoning: Important single events occur that we do not know how to describe mathematically but we can rationalize. This circumstance limits but does not exclude the use of mathematics in economics and finance. For example, we might not have a lot of data on rare events, such as market crashes and depressions, but we can formulate reasonable scenarios for such events that can, in turn, be mathematically represented. Critics of economics and finance as a mathematical science are probably right in saying that these fields cannot be completely represented as unified mathematical theory. To deny that some parts of economic theories can be mathematically described, however, would be unscientific. Finance as an Empirical Science Treating economics and finance more as social sciences is one alternative to prevailing practice. Stricter adherence to the paradigm of empirical science is the other. We will refer generally to this latter approach as “scientific eco- nomics” or “scientific finance.” We can broadly distinguish three main subfields of scientific econom- ics: (1) econometrics and signal processing applied to financial economics, (2) statistical mechanics applied to financial economics, and (3) the theory of complex systems and network theory. Econometrics is the oldest application of scientific principles to economics and finance. It is based on applying statistical methods—in particular, time- series analysis—to empirical data. The diffusion of electronic transactions and the consequent availability of high-frequency and tick-by-tick data have enabled new methods of time-series analysis borrowed from the field of signal processing. Techniques such as econometrics and signal processing can be consid- ered applications of the scientific method in restricted domains in investment management, such as trading and execution. These techniques are based on collecting data, constructing hypothetical models, and then testing the mod- els. The key problem with econometrics and signal processing is the amount
  29. 29. Investment Management 20 ©2014 The CFA Institute Research Foundation of noise in empirical finance data, which makes estimates highly uncertain. The choice of model is rarely based on compelling data. The application of statistical mechanics to financial economics is a new field. Of the results obtained, perhaps the best known is the celebrated presence of “fat tails” in most economic data distributions. The presence of fat tails in distributions implies that large events have a nonnegligible probability of hap- pening. In a Gaussian distribution, on the contrary, large events—say, those more than three standard deviations from the mean of the distribution—can be safely ignored. Not so with fat-tailed, non-Gaussian distributions. Fat tails play a fundamental role in investment management, with important implications for the notion of diversification, risk–return optimization, and risk management. The discovery that economic and financial variables are not Gaussian but exhibit fat tails is a cornerstone of modern financial modeling. The model- ing of fat tails with stable distributions and their application to finance is a major innovation. Fifty years ago, Benoit Mandelbrot (1963) provided the first fundamental attack on the assumption that price or return distribu- tions are normally distributed. His empirical evidence, based on various time series of commodity returns and interest rates, strongly rejected normality as a distributional model for asset returns. Instead, Mandelbrot conjectured that financial returns are more appropriately described by a non-normal stable dis- tribution. Supported by the work of Fama (1963a, 1963b), this result led to a consolidation of the hypothesis that asset returns can be better described as a stable Paretian distribution. Svetlozar Rachev, professor at Stony Brook University, New York; Christian Menn of the University of Applied Sciences in Mainz, Germany; and Frank Fabozzi, co-author of this book, outlined the disruptive impact of stable distributions on financial modeling in their book Fat-Tailed and Skewed Asset Return Distributions (2005). As they noted, the findings of Mandelbrot and Fama caused considerable concern in the finance profession. The authors quoted Paul Cootner (1964), a highly regarded financial economist who taught at both MIT and Stanford, who noted that if financial and economic variables were confirmed to follow a stable distribution, “almost without exception, past econometric work is meaningless” (p. 337). Cootner went on to warn that before the Paretian hypothesis about asset returns should be accepted, more evidence was needed. As Rachev et al. (2005) noted, however, although a preponderance of empirical evidence was against normal distribution and supported fat-tailed distributions of financial variables, the “normality” assumption remained the cornerstone of many leading theories used in finance. In fact, the authors argued, the highly innovative nature of describing financial variables with stable distributions led to a rejection of these distributions, often on the basis
  30. 30. Finance Theory ©2014 The CFA Institute Research Foundation 21 of very weak arguments. For example, a strong objection was that there is no closed-form representation of stable distributions, an objection that the diffu- sion of powerful computers and numerical methods has made obsolete. In addition to the fat-tailed nature of financial phenomena, any empirically based model must take into consideration the fundamental self-referentiality of financial markets and the models we use. Professor Lleo commented, Any model, field, or theory has a sociological dimension. Models are reflec- tive, in the sense that a model that is widely adopted will tend to perform better and better, which in turn speeds up its adoption. This feedback loop can also turn against the model, as we have seen with value at risk [VaR] during the crisis: If all market participants adopt the same set of standards then they will tend to behave homogenously, which speeds up the growth of a bubble and precipitates market crises. Why Are Mainstream Economic and Financial Economic Theories So Resilient? Despite the failings in practical applications and numerous studies that show how unrealistic the assumptions are, mainstream economic and finance theories are remarkably resilient. One explanation is that general equilibrium theories embody the notion of economic rationality. From the point of view of econo- mists, rationality has many advantages. It allows the creation of a sophisticated theoretical construction even when data are missing or difficult to interpret. Benjamin Friedman (2010) professor of political economy at Harvard University, wrote, It sometimes seems that many economists write, and teach, not about the world in which we live but rather the world in which they wished we lived— perhaps because the alternative world is analytically easier to handle, or perhaps because they find the policy implications that would follow in that world more to their liking, or perhaps for yet other reasons. This path is very seductive. Especially in the intellectual arena, few ideas offer more appeal than a model that is simple, elegant, and wrong. (p. 3 of electronic document) MIT’s Professor Lo (2012) suggests that economists suffer from theory envy; that is, their objective is to create a structure that is on a par with their colleagues in the physical sciences. In commenting on the “exalted role of the- ory in economics,” Professor Lo wrote, “Theoretical foundations have become a hallmark of economics, making it unique among the social sciences, but any virtue can become a vice when taken to the extreme of theory envy.” (p. 45) Neoclassical economists defend general equilibrium as an idealized frame- work that represents an economy without the imperfections of real economies; that is, the model is correct and reality is wrong. In finance theory, which has
  31. 31. Investment Management 22 ©2014 The CFA Institute Research Foundation adopted the principles of classical economics, including general equilibrium, the real-world behavior of prices is said to present “price anomalies.” However, another powerful motivation exists: Economic rationality includes faith in the optimality of markets and their self-correcting capabil- ity. In his History of Economics, John Kenneth Galbraith (1987) remarked that economic theory reflects the ideology of the dominating power. For example, the Iron Law of Wages was described in the early nineteenth century by the English banker and economist David Ricardo. Ricardo considered that wages “naturally” tended toward a minimum level—the price that would allow labor- ers to subsist and perpetuate without increase or diminution in their number. For factory owners in an industrializing Great Britain, the idea was quite attrac- tive. Unfortunately, for those same factory owners, Ricardo’s “labor theory of value,” as he called it, also influenced Marx in his early pessimistic views about the possibility that workers might benefit from capitalism. The rest is history. The fact that theory reflects the interests of the dominating power is not limited to economics. In Renaissance France, as the power of French kings was being consolidated, French jurist and political philosopher Jean Boudin put forward a theory of sovereignty that argued in favor of absolutism as the best political system. Thomas Kuhn (1962) analyzed the path through which science makes progress. According to Kuhn’s classical analysis, science starts with the accu- mulation of data and empirical evidence. The tendency is always to defend current theories, grudgingly making adjustments when the theory is no lon- ger tenable, but the accumulation of new empirical evidence can force a para- digm shift that results in new competing theories. Kuhn observed that science, like economic and political theory, is not neutral: Political and ideological influences shape its development. Of several well-known examples, an often cited one comes from the Soviet Union, where the ideas of the biologist and agronomist Trofim Lysenko were imposed in the Stalin era even though they were plainly wrong. Lysenko rejected Mendelian genetics in favor of the hybridization theories of the Russian horticulturist Ivan V. Michurin.7 Lysenko argued that crops’ inheritance was environmen- tally acquired. Scientific dissent from Lysenko’s theory was formally out- lawed in 1948. As a result, Soviet research in biology came to a virtual halt 7 Gregor Mendel was a central European monk and teacher of mathematics, physics, and Greek. He used the microscope to conduct research on the basic facts of heredity. In his research on the common pea plant, Mendel discovered that certain traits show up in offspring without any blending of parent characteristics. The mechanisms of heredity that he discov- ered working on plants are basically the same for all complex forms of life. Michurin was one of the founding fathers of scientific agricultural selection. He worked on hybridization of plants of similar and different origins. The most important problems elaborated by him were intervarietal and distant hybridization.
  32. 32. Finance Theory ©2014 The CFA Institute Research Foundation 23 and programs to improve agricultural output fell far short of their objectives. After 1965, when Lysenko lost all political support, official sanction was bestowed on the view that Michurin was a breeder of genius whose unusual methods can be explained by genetics. Following a series of economic and financial crises that have made it difficult to maintain intact mainstream theories of equilibrium and rational agents, economic and finance theory also might be moving toward a turning point. Many are now calling for modification of the prevailing paradigm or even a paradigm change. But changing a scientific paradigm is never easy. Max Planck (1949 or 1950), a founder of quantum mechanics, wrote, “A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.” Or, as he put it more succinctly: “Science advances one funeral at a time.” Yet another explanation is that mainstream economists and financial economists dominate the major publications and have created an effective barrier to the publication of ideas critical of or challenging mainstream the- ory. Professor Kay (2012) commented on the difficulty of getting published if one does not adhere to mainstream neoclassical thinking.8 He said, You would be told that your model was theoretically inadequate: It lacked rigour, failed to demonstrate consistency. You might be accused of the cardinal sin of being “ad hoc.” Rigour and consistency are the two most powerful words in economics today. . . . [Consistency and rigour] have undeniable virtues, but for economists they have particular interpretations. Consistency means that any statement about the world must be made in the light of a comprehensive descriptive theory of the world. Rigour means that the only valid claims are logical deductions from specified assump- tions. Consistency is, therefore, an invitation to ideology, rigour an invita- tion to mathematics. (p. 52) Other sources have commented on the difficulty of getting published in major professional publications for anything other than what supports the 8 By “mainstream neoclassical thinking,” we mean so-called freshwater economics based on the theories of bounded rationality, the efficient market hypothesis, and rational expecta- tions. This school of thought is often referred to as “freshwater economics” because its major proponents, including Lucas and Fama, come from universities in or near the Great Lakes region, such as the University of Chicago and Carnegie Mellon. The school of thought based on Keynesian economics places less emphasis on theoretical and model consistency and con- siders examples of irrational behavior interesting and important. This school of thought is often referred to as “saltwater economics” because its major proponents, including Shiller and Lo, come from universities on the east and west coasts of the United States, such as Yale, MIT, and the University of California, Berkeley. A new synthesis of the two is referred to as “brackish-water economics.”
  33. 33. Investment Management 24 ©2014 The CFA Institute Research Foundation prevailing economic and finance theory. Bruce Jacobs, principal of Jacobs Levy Equity Management, said: Conflicts of interest in the rarified world of professional publications may seem like an arcane concern, unlikely to have much influence on the real world. But conflicts of interest can lead to self-referential, closed systems that discourage learning and growth. The more closed a system of thinking becomes, the more defensive it is toward criticism, the tighter it holds onto its beliefs, and the less able it is to recognize its own faults. A positive feed- back system is created, in which only affirmation of already held opinions is permitted. Conflict-of-interest standards can weaken the defenses that protect such a faulty system. Indeed, many researchers wanting to publish findings that poke holes in the prevailing theory cannot get published in major economic and finance journals. Most papers that explore new ideas outside the framework of main- stream neoclassical economic and finance theory have been published in such journals as Nature or Physica.9 What is, perhaps, more disturbing is that mainstream journals reject papers that present empirical results and statistical analyses unless the findings are in line with mainstream theory. The accumulation of empirical results is fundamental, however, for the progress of any empirical science. In the hard sciences, from physics to biology, if reported results do not fit existing theories, the results are first verified by other researchers and, if confirmed, a process of theory revision starts. In economics and financial economics, results that do not fit the theory are often simply ignored or are considered anomalies, mak- ing mainstream theory virtually unassailable and resistant to change. The internet, however, may be changing this situation. At least, such is the (optimistic) view of Andrew Haldane (2012), executive director for finan- cial stability of the Bank of England. He remarked that academia’s way of “keeping score” looked “increasingly antiquated.” He wrote, Journal publication remains the main currency, but it is a devalued currency, at least as a medium of exchange for ideas. Some of the top names in the eco- nomics world have taken to social media and the blogosphere to propagate their ideas. This has the benefit not just of immediacy but reach. It amounts to using those network externalities to academic advantage. (p. 139) In summary, mainstream economics and financial economics are not empirical sciences in the sense that physics and chemistry are: Many of the terms used are meaningless; assumptions are unrealistic; and the theory can- not be validated with empirical tests. Despite this empirical failure, main- stream neoclassical theories remain the prevailing theoretical model. 9 Physica is a journal published by Elsevier consisting of subjournals A through E, of which A and E publish peer-reviewed research on econophysics.
  34. 34. Finance Theory ©2014 The CFA Institute Research Foundation 25 The most recent crisis, however, has allowed critics to gain a hearing; new ideas—either more scientific, in that they are based on empirical data, or on the contrary, arguing that economics and finance should be placed back in the realm of the social sciences—are beginning to be discussed seriously. As Professor Lo (2012) wrote: The recent financial crisis has exposed some serious gaps in our understand- ing of the global economy, and the need to take stock and get our academic house in order has never been greater. This presents us with a precious opportunity to make wholesale changes to our discipline that would other- wise be impossible, so we should delay no longer. (p. 48)
  35. 35. ©2014 The CFA Institute Research Foundation 27 2. The Theory and Practice of Investment Management after the Crisis: Need for Change? In the previous chapter, we explored some of the problems with mainstream or classical finance and economics. Continuing to base our discussion on a review of the literature and conversations with finance professionals in aca- demia and the industry, we now consider whether and how the theory and practice of investment management as taught (and practiced) today needs to be revisited. As discussed in the preceding chapter, current mainstream finance theory is embodied in general equilibrium models. These models are idealized math- ematical representations of an economy and markets populated by rational agents who have perfect knowledge of all possible contingencies now and into the infinite future and who optimize the utility derived from consumption and production. Agents are coordinated by price signals. The capital asset pricing model is the prototype of general equilibrium models. As noted in Chapter 1, even many of the theory’s advocates acknowl- edge that these models are unrealistic (or simplistic) and require additional “pieces.” Real agents do not have rational expectations; they interact and can- not be collapsed into a single representative agent. More serious, perhaps, from the point of view of science, is that gen- eral equilibrium models cannot be estimated from empirical data. In particular, the utility function of the representative agent cannot be esti- mated; that is, general equilibrium models cannot be validated. They offer an idealized representation of financial markets and economies at large that does not take into consideration such fundamental elements as the banking system, liquidity, employment and wages, instabilities due to cascades of interactions, and crises. Work is being done to add some of these and other components to the theory. But many are now questioning whether financial economics can be reduced to a global model, useful as such a model might be. So, do the theory and practice of investment need to be revisited? Didier Sornette, a physicist by training and chair of Entrepreneurial Risks at ETH Zurich, summed up the feeling of many of the individuals we talked to for this study. He said, The crash of 2008 certainly put on the radar screen many of the prob- lems with traditional finance. But so did the LTCM [Long-Term Capital
  36. 36. Investment Management 28 ©2014 The CFA Institute Research Foundation Management] crisis in 1998, and so did many other crises. There is a strong incentive in the business to forget lessons. We will now explore some of those lessons with a bearing on the teaching and practice of investment management, namely: • diversification, • optimization—diversification formalized, • the CAPM and similar models, • the efficient market hypothesis, • risk measurement and risk management, and • crises. For each of these topics, after a brief review of the theoretical framework, we present the various opinions and conclude with proposals for change. Diversification Since the pioneering work of Harry Markowitz (1952), diversification has been a fundamental concept in asset management and asset-pricing theories. The notion of diversification can be traced back to medieval merchants and perhaps to well before the Greco-Roman world.10 The concept is so essential that it has been popularized by the adage: “Don’t put all your eggs in one bas- ket.” In finance, diversification implies that you can obtain the same expected returns but reduce your risk by investing in a portfolio of many assets rather than investing in only one or a few assets. From a statistical point of view, diversification is summarized in two mathematical facts: (1) by appropriately choosing weights—that is, the pro- portion of funds invested in each asset—one can reduce the variance of a portfolio while maintaining unchanged its expected return and (2) the min- imum possible variance of a portfolio is smaller than the variance of any of its components. 10 For example, the rabbinical writings of medieval Judaism (e.g., the Talmud) emphasize diver- sification explicitly. A well-known piece of advice is to keep one-third of one’s fortune in busi- ness, one-third in land, and one-third “at hand.” Some have said that the Greco-Roman world did not have a notion of risk because Greeks and Romans believed that the Gods and the Fates determine human fortunes. This conclusion is questionable in light of the high place reserved for “Prudence” in the hierarchy of Greco-Roman virtues. In fact, in antiquity, Prudence was represented as the two-faced god Janus—one face old, the other young because the Ancients believed that Prudence was acquired by consideration of the past and foresight of the future: Does this sound like risk management? Note that Janus was a remote god considered to have rescued men from savagery. Moreover, the ancient Greeks entered into insurance contracts, which are also mentioned in the earlier (1770 BCE) Code of Hammurabi in what is now Iraq.
  37. 37. The Theory and Practice of Investment Management after the Crisis ©2014 The CFA Institute Research Foundation 29 Because it allows one to reduce variance without affecting returns, diversification has often been described as the “only free lunch in financial markets.”11 If, for example, stocks and their returns are uncorrelated and indi- vidual variances bounded, then the variance of a portfolio can be made arbi- trarily small by increasing the number of stocks. Stock returns are correlated, however, so diversification has lower bounds. In fact, market-wide correlation implies the existence of common factors that affect the entire market.12 This is the celebrated separation between diversifiable risk—that is, risk that can be diversified away—and nondiversifiable risk. These properties are purely statistical facts and are, of course, undisputed. What has been questioned is the applicability of diversification. In fact, in the 2007–09 financial crisis, portfolios that were supposed to be well diversified and, therefore, protected from the risk of large losses actually lost significant value. For example, those invested in the SP 500, which is, in itself, highly diversified (but consisting entirely of equities), would have lost 57% from the market’s peak (9 October 2007) to its bottom (9 March 2009). Doubts have been voiced as to the effectiveness of diversification at every level of aggregation. Evariste Lefeuvre (2012), CIO and chief economist for the Americas at Natixis Global Asset Management, com- mented, “Recent empirical analysis shows that expanding the asset mix to [include more] equity-like assets [as well as equities per se] does not provide the expected benefits of asset allocation (the so-called ‘only free lunch’ in finance)” (p. 17). Nevertheless, defenders of diversification argue that diversification always “works” if we define the opportunity set of asset classes broadly enough. The argument is that it is not economically possible for all asset classes to go down together. When “everybody” sells something, they buy something else and whatever they are buying goes up if we expand sufficiently the asset classes. Critics of this claim argue, however, that in a severe crisis, all production and commercial activities can be impaired and the total value of investable assets can go down. Exhibit 2.1 summarizes the defense and critique of diversification according to our conversations with the industry and academic sources and a review of the literature. 11 Diversification is not really a free lunch if assets are priced as if investors are already diversi- fied. In such a condition, which seems likely to be true, not to diversify is wasteful (throwing away one’s lunch). There is no such thing as a free lunch. Nevertheless, it is sometimes peda- gogically useful to refer to diversification in such terms. 12 This statement is not rigorous and would need appropriate mathematical qualifications. Empirical correlations might appear to be random fluctuations of a random matrix, and cor- relations might affect only some sectors and not the entire market.
  38. 38. Investment Management 30 ©2014 The CFA Institute Research Foundation In other words, the defenders of diversification argue that, although it might occasionally fail because of random fluctuations of market parameters, diversification remains a major component of investment decision making. To be maximally effective, however, diversification requires the stability of parameters, covariances, and expected returns. In practice, these parameters change and limit the effectiveness of diversification. Critics argue that diver- sification is ineffective in many economic states, such as when large market swings or crashes occur. A sound use of diversification would imply, there- fore, forecasting the regime shifts between economic states. Note that this debate is part of a broader debate between the proponents of the “rationality of markets” and its critics. Proponents of market rationality believe that large (negative) market swings or crashes are an expression of the business cycle and markets’ random behavior—that is, local events in which asset value is lost but subsequently (rapidly) recovered. Let’s now consider the various issues and points of view related to diversification. Mainstream economic and finance theory are equilibrium theories. Although, theoretically, fundamentals may change and produce long-lasting recessions and crises, barring significant exogenous events such as wars, the slow change of fundamentals can be foreseen in the classical framework and corrective measures can be implemented. The neoclassical framework does not disregard risk, but risk is exogenous in that framework. The defenders of the neoclassical framework argue that diversification is a sound concept. Occasional failures of diversification are not the expression of structural change but result because correlations are stochastic in nature and, therefore, vary randomly. Barton Waring, former CIO for investment strategy and policy of Barclays Global Investors, now active as a writer and lecturer on those topics, commented that after the 2008 stock market crash, people said that the risk models had failed as correlations went to 1. But, he said, This occasional happenstance that correlations go toward 1 is actually perfectly normal. Consider, for example, a 100-year history of two series with a low correlation, created using simulation methods. So, the overall Exhibit 2.1.  The Defense and Critique of Diversification Defense of Diversification Critique of Diversification Diversification is a sound statistical concept that can be fully applied to protect against diversifiable risk. Loss of effectiveness of diversification is the result of unpredictable random fluctuations. The application of diversification, in itself a sound statistical concept, is limited by changes in the parameters of the economy and by nonlinearity and nonnormality. In some economic states, diversification works well; in others, much less so because most returns are negative. For diversification to work correctly, regime shifts must be predicted and diversification adapted to changing regimes.
  39. 39. The Theory and Practice of Investment Management after the Crisis ©2014 The CFA Institute Research Foundation 31 correlation is perfectly as expected, but if you examined three-year subpe- riods, you would see a great deal of variability [in correlation], with some subperiods approaching a correlation of 1, and some approaching zero. Time-varying correlations over the shorter term are perfectly normal even for series that have a rock-solid longer term correlation relationship. Mr. Waring argues that the risk models did not fail. They probably simply experienced normal correlation variability. In addition to the stochastic nature of the covariance matrix, defenders of diversification point out that diversification works only for idiosyncratic risk. It cannot protect against factor-related risk. They observe that investors have to be realistic about the limits of diversification: It cannot protect against the risk associated with common factors. As the proportion of nondiversifiable risk increases, diversification becomes less effective. Paul Pfleiderer (2012), professor of finance at Stanford Graduate School of Business and cofounder of Quantal International, a supplier of portfolio management systems, wrote in response to critics of the MPT paradigm, One of MPT’s key insights is that while investors need to be compensated to bear risk, not all risks are rewarded. The market does not reward risks that can be “diversified away” by holding a bundle of investments, instead of a single investment. By recognizing that not all risks are rewarded, MPT helped establish the idea that a diversified portfolio can help investors earn a higher return for the same amount of risk. (p. 1 of electronic version) Professor Pfleiderer acknowledged that in times of crisis, increased cor- relations reduce the benefit of diversification. He attributed this phenomenon to the (probably) increased importance of macro factors during a crisis. He maintained, however, that “the increased correlations limit, but do not elimi- nate, diversification’s value” (p. 2 of electronic version). Steven Greiner, director of portfolio risk at FactSet Research Systems, observed, The reason some believe diversification didn’t work during the credit crisis of 2008 was that they didn’t understand that only idiosyncratic risk is diver- sifiable. If 95% of your portfolio risk is systematic, whether you own 30 or 300 securities, when the system goes down, you’re going down with it. Dr. Greiner then pointed to a hypothetical portfolio that is not diversified across asset classes or factors. He concluded, “Active weighting is a very poor way to measure exposures and to achieve diversification.” This discussion is part of the debate on the interplay between dynamic asset allocation and diversification. On the one hand, diversification is a “static” concept: Defenders of diversification argue that by investing in a broad diversified portfolio, an investor is protected. Dynamic asset allocation,